Journal Issue

Fiji: Selected Issues and Statistical Appendix

International Monetary Fund
Published Date:
January 2003
  • ShareShare
Show Summary Details

I. Public Investment and Economic Growth1

A. Introduction

1. In the 2002 budget, the government of Fiji announced that growth-enhancement would be a key pillar of its economic strategy. Specifically, the government established a target of raising the sustainable growth rate of real GDP to 5 percent per year, well above the average of barely 2 percent per year achieved over the past half decade. A key element in the implementation of the strategy is to significantly increase the rate of investment in the economy. In this context, the government set an objective of nearly doubling the share of capital expenditure in total government spending to 30 percent over the medium term.

2. This chapter examines the relationship between investment and long-term growth in Fiji. Section B reviews Fijian data, and indicates that the slowdown in the growth rate since independence has been associated with a decline in investment, particularly by the private sector. Section C reviews empirical research on the link between investment and growth, and presents additional evidence for the linkage in the context of small island economies (SIEs) in which tourism plays a major role. Section D concludes with comments on the application of the results to Fiji.

B. Background

3. Since independence in 1970, Fiji has experienced a decline in the average growth rate of real GDP from over 8 percent per year in the 1970s to under 3 percent in the 1980s and 1990s. As a result, the growth rate fell from slightly above the average for a group of SIEs to below the average (Figure 1a).2

Figure I.1Fiji: GDP and Investment

Source: International Financial Statistics.

4. The decline in the economic growth rate has been associated with a marked decline in the share of investment in GDP in the 1980s (Figure 1b). Private-sector investment fell from around 14 percent of GDP through most of the 1970s to around 4 percent in the 1990s, with a sharp decline especially evident following the 1987 coups. Despite some increase in public-sector investment in the late 1980s, aggregate gross fixed investment fell from close to 20 percent of GDP in the 1970s to a little over 12 percent in the 1990s. As a result, the share of investment in GDP fell from being comparable to those for other SIEs in the 1970s and early 1980s to the lowest in the sample in the 1990s (Figure 2). The low rate of private investment appears to have reflected increased concerns regarding political stability and land leases following the 1987 coups. Other factors identified in a 1999 Reserve Bank of Fiji survey include bureaucratic impediments and delays, and shortages of skilled workers.

Figure I.2.Fiji: Investment, 1965–2001

Source: International Financial Statistics.

C. Empirical Estimation

5. Most of the empirical literature on growth emphasizes the relationship between output growth and capital formation.3 In particular, several studies have presented evidence that there is a sizable improvement in potential growth after an expansion of public capital expenditure. For instance, Easterly and Rebelo (1993), using decade average data from 125 countries, show that a 1 percentage point increase in the ratio of general government investment to GDP raises the potential per capita real GDP growth rate by about 0.04-0.05 percentage points.4Gupta et al (2002) focus on the 1990s experience of the 39 poorest countries in the world.5 Using annual data, they show that a 1 percentage point increase in the ratio of capital expenditure to GDP increases long-run per capita GDP growth by about 0.1 percentage points.

6. One difficulty with the use of annual data is the difficulty of separating the long-term supply-side effect of growth-enhancement from the short-run impact on demand of increased investment spending. In addition, the impact on long-run growth is likely to depend on the country’s stage of development and other country-specific characteristics. Although many of the studies in the growth literature control for initial GDP levels, there are no recent cross-country reviews that consider economies with similar institutional characteristics, size, and degree of development as Fiji. Tourism-led SIEs such as Fiji may respond differently to social infrastructure development. Moreover, being a small and isolated country, Fiji is unlikely to experience an industrial transformation, which is usually considered the only source for rapid growth.6

7. With only limited scope for industrialization, it might be postulated that public capital expenditure would have less impact on growth in SIEs than in countries with greater scope for industrialization. At the same time, however, Fiji’s endowment of natural resources gives it a comparative advantage for minerals, fishing, and forestry. Moreover, in the case of tourism, an increase in value as a resort destination could mean that public capital expenditure may raise the growth rate higher than the world average.

8. To capture the effect of capital expenditure in Fiji, a simple cross-country regression analysis is conducted for a sample of SIEs with significant tourist industries. Specifically, with limited availability of data, 1990–2000 data for nine countries are considered in the analysis.7 The regression equation is as follows:

RYGi = α+βCGi+γTGi+δIYi+εi,

where subscript i denotes each SIE, RYG indicates the five-year (1996-2000) average of the real GDP growth rate,8CG represents the five-year (1991-1995) average of the ratio of capital expenditure to GDP, TG denotes the five-year (1991-1995) average of the ratio of total expenditure to GDP, IY denotes the logarithm of the initial (1990) per capita GDP level in U.S. dollars, and ε is the error term. Estimation results are shown in Table 1.

Table I.1.Regression of GDP Growth Rate on Expenditure 1/
VariablesEstimation without

initial GDP level
Estimation with

initial GDP level
Public capital

Total public

Initial GDP level (in logs)-0.25
Number of observations99
R- square0.670.68

Numbers in parenthesis are t-statistics.

Numbers in parenthesis are t-statistics.

9. Several interesting results emerge. Table 1 indicates that capital expenditure in the SIEs increases long-term growth by a higher magnitude to, but not significantly different from, that found in the Easterly and Rebelo analysis: a sustained increase in public capital expenditure of 1 percent of GDP raises the sustainable growth rate by 0.6 percentage points. Other components of expenditure do not seem to affect the growth rate, as the coefficient for total expenditure is not statistically significantly different from zero. Moreover, the initial level of income does not affect potential growth either, probably because income levels in the nine countries are relatively similar.

10. It should be noted also that, since the data are averaged over five years, a one-year-alone increase in public capital expenditure raises growth by only one-fifth of the estimate for a sustained, five-year increase in investment. In other words, a one-year-alone 1 percentage point increase in public capital expenditure boosts the potential growth rate by about 0.12 percentage points. In addition, since both current and capital expenditure should exhibit similar short-run demand stimulus effects, the finding of an insignificant coefficient on total expenditure suggests that the use of five-year averages for variables in the regression effectively eliminates the influence of the short-run effects of increased investment spending on growth, so that the results primarily reflect the impact of investment on long-term growth.

D. Concluding Remarks

11. The main finding of this chapter is that the effect of public capital expenditure on long-term growth in a sample of SIEs with large tourism sectors does not appear to differ substantially from the average for the much-more-heterogeneous sample used in Easterly and Rebelo (1993). Thus, it is possible to use the specific estimates as a rule of thumb to estimate how much Fiji’s government has to invest to in order to achieve a higher economic growth trajectory. Specifically, our estimates suggest that, ceteris paribus, it would be necessary to increase public capital expenditure to over 30 percent of GDP to approach the government’s long-term real annual GDP growth target of 5 percent.

12. Inevitably, such estimates are subject to many caveats. One of particular concern in the Fijian context is the distinction between public and private investment. Much of the recent discussion of the effects of investment on growth is concerned with the impact of private investment. By contrast, the estimates in this chapter focus only on the public component of the investment rate. No allowance is made for the different effects of private and public investment on sustainable growth, nor for possible crowding-out or crowding-in effects as increased public investment either displaces or encourages private investment. Research that has allowed for both types of investment indicates that traditional government expenditure items tend to crowd-out private investment, while other items, such as transport and communication expenditures, crowd-in private investment in developing countries.9


    EasterlyWilliam and SergioRebelo1993“Fiscal Policy and Economic Growth”Journal of Monetary EconomicsVol. 32 pp.417-458.

    BarroRobert and Jong-WhaLee1994“Losers and Winners in Economic Growth” in Proceedings of the World Bank Annual Conference on Development Economicsed. by MichaelBruno and BorisPleskovic pp.267–297.

    • Search Google Scholar
    • Export Citation

    CollierPaul and Jan WillemGunning1999“Explaining African Economic Performance”,Journal of Economic LiteratureVol. 37 (March) pp.64–111.

    • Crossref
    • Search Google Scholar
    • Export Citation

    DewanEdwin and ShajehanHussein2000“Determinants of Economic Growth” Reserve Bank of Fiji WP/00/07.

    GuptaSanjeevBenedictClementsEmanueleBaldacci and CarlosMulas-Granados2002“Expenditure Composition Fiscal Adjustment and Growth in Low-Income Countries” IMF WP/02/77.

    • Search Google Scholar
    • Export Citation

    HabibAhmed and StephenMiller2000“Crowding-Out and Crowding-In Effects of the Components of Government Spending”Contemporary Economic PolicyVol. 18 (January) pp.124–133.

    • Crossref
    • Search Google Scholar
    • Export Citation

    LucasRobert E.2000“Some Macroeconomics for the 21st Century”Journal of Economic PerspectivesVol. 14No. 1 pp.159–168.

    Reserve Bank of Fiji2000Investment Survey: December1999.

1Prepared by Kenichi Ueda (ext. 36368) and Eric Parrado (ext. 34423).
2The SIE sample consists of Barbados, Cyprus, Dominican Republic, Fiji, Jamaica, Maldives, Mauritius, Seychelles, and the Solomon Islands, selected on the basis of similarities in terms of economic structure, income per capita, and data availability.
3See, among others, Barro and Lee (1994) and, more recently, Collier and Gunning (1999) and Dewan and Hussein (2000).
4Their estimates are about 0.4-0.5. However, since they use decade average data (to avoid the contemporaneous stimulus effect), a 1 percentage point increase in the public capital expenditure ratio in a given year yields only a 0.1 percentage point increase in the 10-year average public capital expenditure ratio. Thus, the impact of a one-year increase is 0.04-0.05 percent.
5ESAF- and PRGF-supported countries during the period 1990–2000.
7The countries are listed in footnote 2.
8Again, because of data limitations, the real GDP growth rate is used, not per capita data, unlike in other studies.
9See, among others, Habib and Miller (2000).

Other Resources Citing This Publication